In Paul Veyne’s Foucault (Paris: Albin Michel, 2008, no English translation yet), “discourse” comes as a third medium layer between the subject’s individual consciousness and society at large. If most philosophical and epistemological systems have the knowing subject as epicenter (e.g. Husserl’s intentional phenomenology, or the Heideggerian Dasein), by contrast the social sciences took the side of society, ignoring the subject as a flawed subjective entity. What’s missing is that in-between “language” that tells us what to do and not do, and what to say and not say: in short, something that poses limits to our thoughts and makes every discovery rare and precious. That’s precisely what Foucault meant by discourse. Thus, for art, for example, there is the individual artist on one side, and society at large on the other, but what makes a work of art possible is that tertium quid that defines what is possible at a certain time in a particular society. Discourse would not, however, be able to concretely materialize within a spacio-temporal terrain without the apparatus (dispositif) that would make its very existence possible. The discourse of law in a particular society would not be possible without the various apparatuses of justice through which discourse would operate: the tribunals, the judges, experts, lawyers, and all the institutions that would make a legal case possible.
What is interesting in Veyne’s study are his historical and sociological digressions. As a professional historian of the Greco-Roman empire, Veyne is primarily interested, through his Foucauldian musings, in an epistemology of historical and sociological knowledge. As the sociologist Jean-Claude Passeron has argued, in a book that is subject to Veyne’s attention, we’re into, as far as the social sciences and humanities are concerned, a non-Popperian world: that is, we cannot think those sciences within the narrow categories of induction, deduction, and scientific verification. Passeron and Veyne opt for the Weberian ideal-type notion as a preliminary starting point, only to subject it to some insightful tweaks. Thus, ideal-types like feudalism, the Byzantine Empire, or the Tokugawa Shogunate, are semi-proper names (semi-noms propres): they denote something, and beyond that something that they denote, they have absolutely no credibility. What those names denote are the referents that are need for any historical and sociological denotation of social reality. The meaning of those semi-proper names are defined by the endless referents that denote it. We can call that process of documentation as one of indexation of social reality which would make the semi-proper name meaningful and understandable. We can additionally bring Garfinkel’s ethnomethodology into the picture here through the indexation that is implicit in the users’ need to routinely index speech in order to make one another’s utterances comprehensible. The Foucauldian discourse, through its various apparatuses, would therefore act as the medium through which the individual subjective actions achieve meaning.
Once we define historical and sociological epistemology along that line, it becomes more clear that various political idiosyncratic positions would only look for what they really are: as pseudo-epistemologies of the life-world. For example, the critique of orientalism falsely poses itself as a critique of western knowledge, while confusing in one go orientalism with colonialism and imperialism: one would act in conjunction with the other, or they would all act inseparably from one another. But that’s more of an unsophisticated and idiosyncratic political stance than a usefully genuine cognitive epistemological theory.
